¹ The component parts of the province as attested on inscriptions are Cappadocia, Galatia, Pisidia, Paphlagonia, Lesser Armenia and Lycaonia. Cf. ILS 263; 8971; AE 1925, 126Google Scholar; AE 1929, 98Google Scholar. Pergamon viii 3. 21 (= Smallwood, 214) gives Isauria in place of Lycaonia.

² For the evidence see Magie, , RRAM ii, pp. 1429-30Google Scholar (nn. 12-14). The most active period appears to have been the legateship of A. Caesennius Gallus (80-82); see the evidence cited below, n. 65.

³ See now Mitford, 160-75.

⁴ The legio VI Ferrata was transferred to Judaea some time before the Jewish revolt (Lifschitz, B., Latomus 19 [1960], 109 ff.Google Scholar), and A. Tineius Rufus, legate at the time of the revolt of 132, was of consular rank (Syme, R., Historia 14 [1965], 342Google Scholar; Eck, , Senatoren, pp. 17–18Google Scholar; Schürer, E., History of the Jewish People [revised ed. Edinburgh, 1973], p. 378Google Scholar). See also Keppie, L.J.F., Latomus 32 (1973), 859-65Google Scholar; Bowersock, G. W., JRS 65 (1975), 184Google Scholar.

⁵ For the early literature, see Magie, , RRAM ii, pp. 1435-8Google Scholar n. 22. Two articles by Cumont, Franz are still worth reading: ‘Le Gouvernement de Cappadoce sous les Flaviens’. Bull. Acad, de Belgique 1905, 197–227Google Scholar; and ‘L'Annexion du Pont Polemoniaque et de la Petite Armenie’, Anatolian Studies presented to Sir W. M. Ramsay (Manchester, 1923), pp. 109-19Google Scholar. See, more recently, Eck, , Senatoren, p. 3Google Scholar; Mitford, 166-7. I have obtained some valu-able information from a recent doctoral thesis by Schieber, A. S., The Flavian Eastern Policy (State University of New York at Buffalo, 1975)Google Scholar.

⁶ Tac, . Ann. xiii 8Google Scholar; 35: ‘habiti per Galatiam Cappadociamque dilectus’ (cf. xv 6); cf. Sherk, , Legates, pp. 32 fGoogle Scholar.

⁷ C. Rutilius Gallicus is attested praetorian legatus provinciae Galaticae during the 60s (Smallwood, E.M., Documents illustrating the Principates of Gaius, Claudius & Nero [Cambridge, 1967], 244Google Scholar). According to Statius (Silvae i 4. 76-7Google Scholar) he held office for nine years. Cf. Syme, , Klio 30 (1937), 230CrossRefGoogle Scholar; Magie, , RRAM ii, p. 1412Google Scholar n. 40; Sherk, , Legates, pp. 35-6Google Scholar.

⁸ Tac, . Ann. xiv 26Google Scholar (for the text see below, n. 70); Suet, . Nero 18Google Scholar. Cf. Cumont, ‘L'Annexion …… (note 5 above), 112.

⁹ Reinach, 133-57, esp. 13947; cf. PIR ², A 1052Google Scholar.

¹⁰ Tac, . Ann. xiii 7Google Scholar; xiv 26; cf. xv 25: ‘tetrarchis ac regibus…iussisCorbulonisobsequi’.

¹¹ In 63 Corbulo had commanded VI Ferrata, III Gallica, V Macedonica and XV Apollin-aris (Tac, . Ann. xv 26Google Scholar; cf. ILS 232= Smallwood, 51 b). Of these legions V Macedonica and XV Apollinaris joined the army of Vespasian in Judaea in winter 66/67 (Jos, . B.J. iii 8. 65Google Scholar; Tac, . Hist, v 1Google Scholar). VI Ferrata had been withdrawn before 69 to reinforce the army of Syria (Tac, . Hist, ii 83Google Scholar; cf. i 10). As for III Gallica it had first been transferred to Syria and then moved to Moesia by the early months of 69 (Tac, . Hist, ii 74Google Scholar; cf. Ritterling, , RE XII 1521)Google Scholar.

¹² Jos, . B.J. ii 366-7Google Scholar; cf. Tac, . Hist, ii 83Google Scholar.

¹³ ILS 8904 = MW 86. For the importance of the road, see Mitford, 165-6.

¹⁴ Jos, . B.J. vii 18Google Scholar. For the legion's later history, see Ritterling, , RE XII 170Google Scholar; Magie, , RRAM ii, p. 1436 n. 22.Google Scholar

¹⁵ Eck, , Senatoren, p. 3Google Scholar; he observes correctly (n. 6): ‘Natürlich besagt die Schaffung einer von einem Legaten verwalteten Provinz noch nichts iiber die Einfiihrung des grossen Piovinzkomplexes’.

¹⁶ . For the chequered history of XII Fulminata under Nero see Tac, . Ann. xv 6Google Scholar; 10; 26; Jos, . B.J. ii 500 ffGoogle Scholar.

¹⁷ IRT 346 = MW 303.

¹⁸ Eck, W., ‘Die Legaten von Lykien und Pamphylien unter Vespasian’, ZPE 6 (1970), 65–75,Google Scholar has argued that Lycia lost its liberty in the latter years of Nero, on the strength of two enigmatic inscriptions honouring Sex. Marcius Priscus (TAM ii 396Google Scholar; 131). But the theory involves rejecting the explicit statement of Suetonius, Vesp. 8.4, that it was under Vespasian that Lycia was reduced in provinciarum formam (cf. Jones, C.P., Gnomon 45 [1973], 690 f.Google Scholar), and Eck's suggestion (71, cf. Senatoren, p. 4) that Vespasian deprived Lycia not of libertas but of particular privileges granted by his predecessors is, to say the least, not compelling. It seems clear, however, that Lycia was under direct Roman government from the opening years of Vespasian's reign, and it was not immediately associated with Pamphylia.

¹⁹ He is attested both at Balburae in Lycia (IGR iii 466Google Scholar) and at Attaleia in Pamphylia (SEG vi 648Google Scholar). For discussion see Eck, , ZPE 6 (1970), 72-3Google Scholar.

²⁰ Tac, . Hist. ii 9Google Scholar; the full nomenclature is supplied by IRT 346.

²¹ AE 1967, 492 (cf. Levick, B. M., Anatolian Studies 17 [1967], 101ff.CrossRefGoogle Scholar): ‘[Imp. Caesari Vespasiano Aug./ p]ontif[ici maximo/ t]ribunic[iae potes/tjatis cos [11/ L. Nonius] Calpujrnius/ L.f. Pomp(tina) Asprenas]’. The identification is tentatively accepted by Eck, , Senatoren, p. 114 n. 16Google Scholar.

²² Reinach, Th., REA 16 (1914), 144 ffCrossRefGoogle Scholar. Reinach (147) omitted to count inclusively and assumed that Aristobulus' seventeenth regnal year was 71/2. Magie, , RRAM ii, p. 1435Google Scholar n. 21, gives the correct date: 70/1.

²³ Jos, . A.J., xx 158Google Scholar; Tac, . Ann. xiii 7.Google Scholar

²⁴ Reinach, 133; 149-53, describes the Nicopolis coins. They are clearly dated to the 16th year of Trajan (TO FI), that is, 25 Jan. 113-114. The year of the city of Nicopolis is given as 42. The coin struck by the is equally explicitly dated to Trajan's 17th year (TO ZI), and it was clearly issued to celebrate the annexation of Armenia (Pick, B., REA 16 [1914], 283-9CrossRefGoogle Scholar). The year of the is given as 43. The two issues are consistent and indicate that the era of the province began in 72. Mitford, 116 n. 35, claims that the coins do not exclude late 71 (so apparently Magie, , RRAM ii, p. 1435Google Scholar n. 21); but, once again, if we reckon inclusively, the era cannot have begun before 72. 25 See the excellent description by Mitford, 165-6.

²⁵ See the excellent description by Mitford, 165-6.

²⁶ For the XVI Flavia Firma see Dio lv 24. 3; Ritterling, , RE XII 1765Google Scholar. Its presence at Satala is now confirmed by inscriptions on the spot, which record a signifer, C. Quintianus Maximus, and a miles, C. Trebon[ius] (Mitford, 165 f.; no. 3).

²⁷ Jos, . B.J. vii 226Google Scholar. This Aristobulus, who held the kingdom of ‘Chalcidice’ at the time of the annexation of Commagene, has been identified as Aristobulus of Lesser Armenia (Wilcken, , RE II 910Google Scholar; Reinach, 141). The equation, though attractive, is by no means secure.

²⁸ Syme, , Tacitus, p. 31Google Scholar n. 1. The suggestion has been accepted by Eck, , Senatoren, pp. 3Google Scholar n. 7; 115 ff.; and by Bowersock, G. W., JRS 63 (1973), 134Google Scholar f.; 140. Mitford, 173 n. 90, expresses scepticism. The argument would be settled if L. Vidman's redating of Traianus' consulship to 72 were accepted (Listy Filologické 98 [1975], 66–71Google Scholar). But that redating depends on the juxtaposition of the controversial Fragment XI of the Fasti Ostienses with two recently discovered fragments, which seem to date to 72. The fragments may indeed come from the same general area of the stone (Vidman, 67), but there seems no compelling reason to ascribe them to the same year. Fragment XI may still belong to the year 70 and with it the consulship of Traianus.

²⁹ Cf. Magie, , RRAM i, p. 575Google Scholar: ‘It seems probable that this attack aroused in the Roman government the fear of a similar inroad into Lesser Armenia or Cappadocia.’ The idea that the Transcaucasian Sarmatians posed a threat to the Roman east goes back to Mommsen, The Provinces of the Roman Empire (London, 1909), Vol. ii, pp. 61 fGoogle Scholar.

³⁰ Dio lxix 15. 1. For full discussion of this episode see my forthcoming article ‘Arrian and the Alani’, to appear in HSCP 81 (1977)Google Scholar.

³¹ Dio significantly states that the Alani touched upon Armenia and Cappadocia. We can infer from Arrian's disposition (Ectaxis 7; 14-15) that the area of hostilities was Lesser Armenia.

³² Strabo xi 5. 8 (596). For details see Magie, , RRAM i, pp. 407-12Google Scholar.

³³ ILS 986 = MW 261; cf. Pippidi, D. M., Epigraphische Beiträge zur Geschichte Histrias (Berlin, 1962), pp. 121 ffGoogle Scholar.

³⁴ Jos, . B.J. vii 89–95Google Scholar.

³⁵ For a good exposition of the theory of the ‘Sarmatian menace’, see Anderson, J. G. C., CAH x 775-8Google Scholar. See also Magie, , RRAM i, pp. 574 f.Google Scholar; Sherk, , Legates, p. 40Google Scholar.

³⁶ Tac, . Ann. vi 33Google Scholar: ‘Sarmatas, quorum sceptuchi utrimque donis acceptis more gentico diversa induere’. Cf. Ann. xii 15–16Google Scholar (the Aorsi and Siraces take different sides, for and against the Bosporan kingdom).

³⁷ Jos, . A.J. xviii 97Google Scholar (the Alani are specifically mentioned: cf. Tac, . Ann. vi 33-5Google Scholar; Dio lviii 26. 3-4).

³⁸ So vonGutschmid, A., Geschichte Irons (Tübingen, 1885), p. 133Google Scholar; Mommsen, , Provinces of the Roman Empire ii, p. 62Google Scholar n. 1; Anderson, A. R., TAPA 59 (1928), 146-8Google Scholar; Carrata-Thomes, F., Gli Alani nella politico orientale di Antonino Pio (Univ. di Torino, Publ. d. Fac. d. Lett. e Filos. 10 [1958]), p. 13Google Scholar.

³⁹ On the Caspian Gates see von Stahl, A. F., Geogr. Journal 64 (1924), 318-19Google Scholar; Treidler, H., RE XXII 322-33Google Scholar; Standish, J. F., Greece & Rome 17 (1970), 17–24CrossRefGoogle Scholar.

⁴⁰ Pliny, , NH vi 30Google Scholar: ‘Portae Caucasiae magno errore multis Caspiae dictae’; cf. vi 40 : ‘sunt autem aliae Caspiis gentibus iunctae, quod dignosci non potest nisi comitatu rerum Alexandri Magni’. Cf. Anderson, , TAPA 59 (1928), 136-7Google Scholar, for a list of ancient texts referring to the Darial Pass as the Caspian Gates.

⁴¹ Jos, . B.J. vii 246Google Scholar: .

⁴² This was in fact argued by Gutschmid, Geschichte Irons, p. 134; Kleine Schriften (Leipzig, 1892), Vol. iii, pp. 111-13Google Scholar; J. Marquardt went further and suggested that the Hyrcani were yet another confusion in Josephus, being a misunderstanding of the Armenian term for the Iberians (Zeitschr. d. deutsch.– morgend. Ges. 49 [1895], 692Google Scholar). He later sensibly retracted the theory (Caucasica 8 [1931], 78Google Scholar).

⁴³ For the eastern Alans see Ptolemy iii 5. 19; vi 14. 9. They are apparently mentioned in the Annals of the Han dynasty of China, which record that the kingdom of Yen-t'sai (in the vicinity of the Aral Sea) changed its name to A-lan; cf. Enoki, K., Central Asiatic Journal 1 (1955), 47–50Google Scholar. Taubler, E., Klio 9 (1909), 19 ff.CrossRefGoogle Scholar, argued that it was these eastern Alani who invaded Media in 72.

⁴⁴ Dio lxvi 15. 3. It is not stated that the expedition was directed against the Alani, but the parallel passage of Suetonius (Dom. 2. 2) is quite specific. The Parthian appeal is usually dated to 75 (see, most recently, Bowersock, , JRS 63 [1973], 135Google Scholar);but Dio's chronology is not precise. The only episode explicitly dated to Vespasian's sixth consulship is the dedication of the precinct of Pax (lxvi 15. 1). There follows a series of anecdotes, all excerpted by Xiphilinus and all lacking any specific chronological indication. One incid-ent, the capture of the Gallic rebel Sabinus, is known to have occurred in 79 (cf. Tac, . Hist, iv 67Google Scholar). Dio tends to be digressive and biographical in his history of the imperial period – ‘some chronological structure is preserved but it is vague and uncertain’ (Millar, F., Cassius Dio [Oxford, 1964] , p. 71Google Scholar).

⁴⁵ Jos, . B.J. vii 251Google Scholar: .

⁴⁶ Tac, . Hist. iii 47-8Google Scholar.

⁴⁷ In the summer of 69 Mucianus was on his way through Moesia with the other Syrian legion, VI Ferrata, and no less than 13,000 vexillarii (Hist, ii 83Google Scholar). That would have left the remaining legions extremely undermanned.

⁴⁸ Arrian, , Periplus 10.1Google Scholar; 11. 4.

⁴⁹ Jos, . B.J. ii 366Google Scholar, claims that the fleet controlled theHeniochiand Colchians, immediately to the south of the Chobus.

⁵⁰ For the date see Periplus 17. 3, which mentions the death of Cotys II, whose coinage covers the period 123/4 - 131/2 (cf. PIR2 I 276Google Scholar; 516).

⁵¹ Peripl. 11. 1–3Google Scholar; 13. 3.

⁵² Peripl. 11.1–2.Google Scholar

⁵³ Strabo xi 2. 12 (496); cf. Cumont, ‘Annexion du Pont’ (above n. 5), 110-11.

⁵⁴ Cf. Tac, . Ann. xiii 39Google Scholar; the supply route through the mountains of Lesser Armenia was garrisoned by Roman troops.

⁵⁵ Tac, . Ann. xv 9Google Scholar.

⁵⁶ Jos, . B.J. ii 366-7Google Scholar; cf. von Domaszewski, A., Rh. Mus. 47 (1892), 207-18Google Scholar.

⁵⁷ In the opening months of 70 IV Scythica alone of the eastern legions was available for action. The three Jewish legions with XII Fulminata and vexillationes from Egypt were fully engaged outside Jerusalem (Tac, . Hist. v 1Google Scholar; Jos, . B.J. v 46Google Scholar). III Gallica was dispatched back to Syria from Rome on 1 January 70 (Tac, . Hist, iv 39Google Scholar) but the journey would have taken several months. As for VI Ferrata, it was almost certainly detained in Moesia by the Sarmatian invasion. It is not attested in Syria before the invasion of Commagene in 72/3 (Jos, . B.J. vii 225Google Scholar). Even Syria was not tranquil in 70, with unrest between the Greek and Jewish communities of Antioch (Jos, . B.J. vii 42–62Google Scholar). That may well have been enough to keep the legate of IV Scythica occupied (cf. vii 58-60).

⁵⁸ For these garrisons see Arr, . Periplus 3.1Google Scholar; 6.1-7 (cf. ILS 2660); 9.3; 10.3. Trapezus seems to have been occupied by vexillationes from the regular Cappadocian legions (CIL iii 6745Google Scholar; Mitford, 163-4); they are not mentioned in the Periplus, but the omission is prob-ably fortuitous. Arrian is concerned primarily with the statues of the emperor (Peripl. 1-2), and any details of the garrison force may have been reserved for his official Latin report (Peripl. 6.2Google Scholar; 10.1). In 69 the auxiliary cohort stationed at Trapezus had been massacred (Tac, . Hist, iii 47Google Scholar) and a regular garrison of legionary troops may have been thought necessary.

⁵⁹ Cumont, F., Bull. Acad, de Belgique 1905, 217Google Scholar; cf. ‘L'Annexion du Pont’, 113 f. See now Kreiler, B., Die Statthalter Kleinasiens unter den Flaviern (diss. Munich, 1975), pp. 75 fGoogle Scholar.

⁶⁰ Cf. Magie, , RRAM i, p. 574Google Scholar: ‘It is also unlikely that mere raids by brigands are meant, for these could have been repelled by the Cappadocian militia’. In 69, however, the militia had been found sadly wanting and precious legionary troops had to be diverted from Syria under Virdius Geminus (Tac, . Hist, iii 48Google Scholar).

⁶¹ Cic, . Rep. iii 35Google Scholar: ‘noster autem populus sociis defendendis terrarum jam omnium potitus est’. Cf. de Off. ii 26. For an excellent analysis of Augustus' ‘defensive aggression’ see Brunt, P.A., JRS 53 (1963), 170-6Google Scholar. Marcus Aurelius apparently continued to observe fetial procedure during his northern wars (Dio lxxi 33. 3: .

⁶² Jos, . B.J. vii 220-5Google Scholar. For Antiochus' previous services, see Tac, . Ann. xiii 7Google Scholar; 37; xiv 26; Jos, . B.J. ii 500Google Scholar; iii 68; v 461; Tac, . Hist, ii 81Google Scholar; v 1. His son, Epiphanes, twice distinguished himself in action for the Romans (Tac, . Hist, ii 25Google Scholar; Jos. B.J. v 460-5Google Scholar).

⁶³ Jos, . B.J. vii 238-43Google Scholar. The family later moved to Athens where the king's grandson, C. Iulius Antiochus Philopappus, enjoyed citizenship and held the archonship. Trajan adlected him to praetorian rank, and he was suffect consul in May 109 (cf. PIR², 1151).

⁶⁴ The first known praetorian legate is Ti. Iulius Celsus Polemaeanus, who held office around 79/80 under M. Hirtius Fronto Neratius Pansa (ILS 8971 = MW 316). His successor was L. Iulius Proculeianus, attested legate under Titus while A. Caesennius Gallus was governor (AE 1964, 4Google Scholar [Comana]; cf. Eck, , Senatoren, p. 3Google Scholar n. 9). C. Antius A. Iulius Quadratus was legate during the early years of Domitian's reign IGR iv 384Google Scholar = Smallwood, 213; AE 1929, 98Google Scholar).

⁶⁵ ILS 263 = MW 105. His activity is specifically attested on five milestones dated between 80/81 and 82/3 (Mitford, 172 n. 87), as well as a dedication from the fort of Dascusa on the Euphrates (Mitford, 172-3).

⁶⁶ For the geography of the Caucasian kingdoms see Strabo xi 3. 1 (499) – 4. 8 (503). The article by Treidler, H., RE Suppl. ix 1899–1911Google Scholar, is unsatisfactory in its coverage. There are important, if diffuse, observations made in passing by Grosso, F., ‘Aspetti Sella politica orientate di Domiziano’, Epigraphica 16 (1954), 117-79.Google Scholar

⁶⁷ For Pompey's campaign see Magie, , RRAM i, pp. 358-9Google Scholar; ii pp. 1225-7 nn. 13-15. Antony's general, P. Canidius Crassus, also won a victory over the two kingdoms in 36 B.C. (Dio xlix 24. 1; Plut, . Ant. 34. 10Google Scholar; Strabo xi 3-5 [501]). Both are named in Augustus' list of embassies seeking amicitia (Res Gestae 31. 2).

⁶⁸ 0n this episode see Debevoise, N. C., A Political History of Parthia (Chicago, 1938), pp. 174-8Google Scholar; Magie, . RRAM, i pp. 551-3Google Scholar.

⁶⁹ Tac, . Ann. xiii 37.Google Scholar

⁷⁰ Tac, . Ann. xiv 23Google Scholar fin. – invasion by the Iberians of the territory of the Mardi, to the north-east of Lake Van (cf. Ptol. v 13. 10; Strabo xi 13. 3 [523]). The two peoples were neighbours and perhaps hereditary enemies. See also Tac, . Ann. xiv 26Google Scholar, with Gronovius' emendations of the manuscript reading (Pharasmani Pole <jnoni> que for pars nipulique).

⁷¹ Tac, . Ann. xiii 41Google Scholar: ‘longinquum regis iter et Medi an Albani peterentur incertum’.

⁷² Suet, . Nero 19. 2Google Scholar; Dio 1v 24. 2. For the early history of the legion see Ritterling, RE XII 1407-10Google Scholar.

⁷³ Tac, . Hist, i 6Google Scholar – the force had already been sent out (praemissos) and Nero had been obliged to recall it (revocaverat).

⁷⁴ Pliny, , NH vi 40Google Scholar: ‘vix ullo propter oppositos montes aditu ad Caspium mare’.

⁷⁵ Mommsen, , Provinces of the Roman Empire, Vol. ii, p. 62Google Scholar n. 1. Though Mommsen's emendation has not been widely accepted, there has been general agreement that the Sarmatians north of the Caucasus were the principal object of Nero's expeditions: cf. Täubler, , Klio 9 (1908), 14Google Scholar; Anderson, , CAH x 777-8Google Scholar; Carrata-Thomes, op. cit. (n. 38 above), p. 12; Pippidi, , Epigraphische Beiträge, pp. 131Google Scholar f. Schur, W., Die Orientpolitikdes Kaisers Nero (Klio Beih. xv 1923), pp. 62Google Scholar f., evolved a famous hypothesis that Nero intended to encircle Armenia with client states so as to control the trade routes leading to Bactria and India. Not surprisingly it had little success (Magie, , RRAM ii, p. 1418Google Scholar n. 63).

⁷⁶ ILS 986. 24-5Google Scholar: ‘Scytharum rege{m} a Cherronensi quae est ultra Borustenen opsidione summoto’. The only Sarmatian people known to have attacked the Bosporan kingdom are the Siraces (Tac, . Ann. xii 15. 2Google Scholar; cf. Latyschev, , IPE ii 423Google Scholar, A.D. 193). The neighbouring Scyths were a more serious and consistent nuisance.

⁷⁷ Compare Jos, . A.J. xviii 97Google Scholar, where the Ambrosian codex alone reads The manuscript consensus is for

⁷⁸ 78 For the strategic importance of Harmozica see Strabo xi 3.5 (501); Pliny, , NH vi 29–30.Google Scholar

⁷⁹ ILS 8795; IGR iii 133Google Scholar; MW 237;SEG 20 (1964), 112Google Scholar. See most recently Boultonova, A. I., ‘Quelques notes sur l'inscription de Vespasien trouvée à Mtsketha’, Klio 53 (1971), 213-22Google Scholar. Problems remain concerning the identity of the ‘Iamasaspus’ who is associated with Mithridates on the stone. Boultonova suggests (215 ff.) most implausibly that he is to be identified with Amasaspus, the Iberian prince who died fighting for Trajan at Nisibis in 115 (IGR i 192Google Scholar = IG xiv 1374Google Scholar). At present the most cogent interpretation is that of Tseretheli, G., VDI 1960, 123-33Google Scholar, who reads: and interprets the form as a Greek transcription of the Armenian form Iamazaspuhi. In other words, the inscription gives both the father and the mother of Mithridates. That explains the fact that the concluding epithets are in the singular.

⁸⁰ AE 1951, 263; cf. Yampolsky, Z., VDI 31 (1950), 177 f,Google Scholar; Grosso, F., Epigraphica 16 (1954), 117 ffGoogle Scholar. The inscription is usually omitted in general histories; Syme, , Tacitus, p. 248Google Scholar n. 2, and Mitford, 167 n. 42, are honourable exceptions.

⁸¹ Yampolsky merely says ‘at 70 kilometres from the town of Baku’. Grosso discusses the possibilities for the site and comes to the conclusion that the area of Kilyazi is the best location, ‘dove le propaggini del Caucaso sono piu prossimi alia costa’ (Epigraphica 16 [1954], 124Google Scholar n. 1).

⁸² Statius, , Silv. iv 4. 63-4Google Scholar, discussing the areas where Vitorius Marcellus might serve as legionary legate after his praetorship.

⁸³ The fragmentary career inscription of M. Hirtius Neratius Pansa refers to a post as leg. pr. pr. [‐‐‐‐exercit] us qui in A [‐‐‐‐], which has been interpreted as command of an eastern expedition; in A[rmeniam Minorem] or in A[lanos] or in A[lbanos missus est ‐‐‐] (Torelli, M., JRS 58 [1968], 173Google Scholar). The inscription is clearly too fragmentary to permit certain conclusions (the evidence seems insufficient to justify Torelli's exclusion of the formula qui in ‐‐‐‐ est, and no major campaign is likely to have been placed in the hands of a man of praetorian rank).

⁸⁴ For Vespasian's imperatorial salutations during these years see Weynand, , RE VI 2655-69.Google Scholar He was imp. ix in late 72 (ILS 246), imp. x in early 73, while still censor designate (CIL ii 5217Google Scholar; v 4312); imp. xi in early 74 (CIL ii 2322Google Scholar) and imp. xiv by 21 May 74 (CIL xvi 20Google Scholar). For the prolongation of the imp. xiv titulature into 76 see ILS 8904 = MW 86.

⁸⁵ Strabo xi 3. 4 (500).

⁸⁶ Arrian, , Peripl. 11.2Google Scholar. The garrison of five cohorts at the mouth of the river Apsarus was probably established under Hadrian as a defence against Iberia (Arr, . Peripl. 4. 1–2Google Scholar; cf. ILS 2660 = Smallwood, 308).

⁸⁷ Dio 1xviii 19. 2. See also Arr, . Peripl. 11. 2–3Google Scholar; with Magie, , RRAM ii, p. 1465Google Scholar n. 32. The site of the conference is given as Satala in the Excerpta de Legationibus, but as Elegeia in Xiphilinus' epitome.

⁸⁸ Albanis regem dedit: Festus, , Brev. 20. 2Google Scholar; cf. Eutropius viii 2. 2; Jerome, , Chron. p. 194Google Scholar (Helm). The episode is commemorated on the gold and silver issues struck by Trajan with the legend REGNA ADSIGNATA (Strack, P. L., Untersuchungen zur röm. Reichsprägung des zweiten Jahrhunderts, Vol. i [Stuttgart, 1931] , pp. 222 f.)Google Scholar.

⁸⁹ Themistius, , Orat. 34. 8Google Scholar (T. 13 Roos).

⁹⁰ SEG xv 836Google Scholar = MW 238.

⁹¹ Victor, Aur., de Caes. 9. 10Google Scholar; Anon, . Epit. de Caes. 9. 12.Google Scholar

⁹² For the ornamenta triumphalia see Mitet i 5 (1919), p. 53Google Scholar = MW 263; Bowersock, , JRS 63 (1973), 134Google Scholar. For the Parthica laurus see Pliny, . Pan. 14. 1Google Scholar; 16. 1.

⁹³ Cf. Syme, R., Tacitus, pp. 30-1Google Scholar; Bowersock, op. cit. 134-5.

⁹⁴ Aurelius Victor and the Epitome de Caesaribus (cited above) differ whether Vologaeses was forced to conclude peace by a war or metu solo. The text of Victor responds more easily to emendation; Conn's 〈sine〉 bello immediately suggests itself (cf. Magie, , RRAM ii, p. 1438Google Scholar n. 23). Bowersock, 138, has argued that an inscription on the north-west gate of Gerasa refers to a ; if the supplement is correct, that would seem decisive.

⁹⁵ See the argument of Bowersock, 135, who unfortunately accepts Dio's alleged date of 75 for the Parthian appeal (see n. 44 above). He therefore suggests that the threat of war came two years earlier in the aftermath of the annexation of Commagene. But the annexation of Commagene was precisely the period of the Alani invasion (Jos, . B.J. vii 244Google Scholar, placed between the expedition against Commagene and the beginning of Flavius Silva Nonius Bassus' expedition against Masada), the period of greatest Parthian weakness. We should therefore date the threat of war later and place the appeal to Rome around 72.

⁹⁶ Vespasian is still imp. xiv on the Melik Serif inscription of early 76 (ILS 8904 = MW 86). By 2 Dec. 76 he was imp. xviii (CIL xvi 21Google Scholar = MW 400). Cf. Weynand, , RE VI 2669: ‘eine Zahl ist sicher durch den Krieg des Trajan veranlasst’.Google Scholar

⁹⁷ Tac, . Ann. xv 17. 3Google Scholar; Dio 1xii 22. 2-4.

⁹⁸ This paper was first delivered at a Research Seminar in Ancient History held at the University of Tasmania in February 1976. I am grateful for remarks made by the audience on that occasion and to Professor G. W. Bowersock in particular for his valuable comments on the manuscript. The errors which remain are of course my own.